Industrial agriculture and rural community degradation.
MacCannell, Dean

In Swanson, L.E. (ed.) Agriculture and Community Change in the U.S.: The Congressional Research Reports. pp.15-75, 325-355. Westview Press, Boulder, CO. 1988

Reviewers note: This study, done for the US. Office of Technology Assessmen4 looks at the social conditions in rural communities in four Sunbelt states: California, Arizona, Texas and Florida. It was conducted by the UC Davis Macrosocial Accounting Project, following up on earlier work in California (see references) and examines, on a statistical level the negative relationship between the trend toward increasing farm size and the social conditions in rural communities. This study uses macrosocial accounting (MSA) methods for data collection and analyses which are similar to methods used in agricultural economics. However, instead of modeling costs and benefits of different policies and practices at the level of the firm, MSA describes regional social structure. In this study, MSA is used to examine the relationship between agricultural variables and rural community conditions to determine the social costs and benefits of differences in farm structure.

MacCannell draws two primary conclusions from this study on the social conditions in rural communities of four Sunbelt states-- California, Arizona, Texas and Florida: 1) An advanced, industrial-type agriculture is now well established in the U.S. Sunbelt; and 2) Evidence suggests that this new form of agriculture is associated with substantial deterioration of human living conditions in nearby rural communities. These conclusions challenge the assumptions that have previously informed agricultural and rural policy, i.e. that agricultural economic development generally benefits workers and communities. Acceptance of MacCannell's conclusions will require rethinking policies that have encouraged the dominance of an industrial agriculture in the Sunbelt and/or those policies that will address the social and environmental damage associated with this industrial type of agriculture.

Development of Industrial Agriculture in the Sunbelt

To put this study into perspective, MacCannell reviews the development of industrial agriculture in the Sunbelt and its relation to the existence of the family farm. Despite global industrialization, the family farm continues to be the dominant form of agriculture worldwide, and persists in many parts of the U.S. and California. It was only after World War II that a unique set of circumstances in the U.S. Sunbelt developed and led to the establishment of industrial agriculture on a regional basis. These factors include:

• A historical pattern of large sized farms
  
• State/Federal water projects and subsidized irrigation
  
• Low paid, unorganized foreign labor
  
• Harvest mechanization and other uses of high technology

The confluence of these preconditions allowed industrial agriculture to flourish in the U.S. Sunbelt. The largest farm operators in this region were able to exploit their natural, historical and political advantages by combining government support programs, irrigation systems, foreign labor and new technologies to become preeminent in the national and global agricultural economies. Conventional wisdom suggests that the rapid pace of economic development which occurred in this agricultural sector should improve the economic and social conditions in surrounding rural communities. Yet, this study shows just the opposite. "...it is in exactly those areas where farming is the most modern, rational and economically profitable that the worst general social conditions are found." (p. 17) Although industrialization has brought benefits in terms of better technology and economic profitability to farm owners, it is associated with poverty for the workers and the community. For example, in the most productive agricultural tracts in the study area, poverty rates were from 5 to 40 percentage points higher than in other rural areas of the U.S.

Methodology

Agricultural counties in the four Sunbelt states are the units used in this MSA analysis. Ninety-eight counties were selected based on their ranking in the top 100 counties in agricultural sales nationwide or on their ratio of agricultural sales to population (at least $2000 a year per capita). Most of these counties were dominated by agriculture and mainly rural. Data from these counties were used for descriptive statements as well as for regression analyses of the relationship of industrial farm structure and rural community conditions.

Description of Community Social Conditions

Before describing his regression analyses, MacCannell gives some background descriptive information about social conditions in selected counties in the study area. Some counties and tracts were selected for closer observation if 20 percent of the labor force was classified as "farm laborers" in 1970 and/or 15 percent or more was classified as workers in "farming, fisheries and forestry" (the closest category to farm laborer in the 1980 census) in 1980. Both field observations and tract-level data were gathered for these tracts and counties. The table below summarizes some of MacCannell's findings for California:

As the table shows, housing conditions (using lack of plumbing as an indicator) were worse m the selected counties than the state average. For example, in Imperial County, lack of plumbing is 20 times higher than in the state. The same general pattern emerges from the data on crowding (as measured by percent families living in a house with more than five persons/room). Communities in Fresno and Imperial counties tended to have more families living in crowded conditions compared to the state average.

Average family incomes are 14 percent and 27 percent lower than the California average in Fresno and Imperial counties, respectively. In fact, poverty affects a much larger segment of the population in these areas as the figures show. It is interesting to note that despite the fact that poverty is higher in these counties, employment remains relatively high.

The Regression Analysis

Based on descriptive analyses of rural community conditions in areas of industrialized agriculture (of which the above discussion is just a small part), as well as previous work of the Macrosocial Accounting Project, MacCannell has built his "industrialization-degradation" hypothesis. This hypothesis asserts that increasing rural community degradation is positively correlated with increasing agricultural industrialization. To test this hypothesis, MacCannell uses regression analyses in which farm structure and agricultural technology indicators are entered as independent variables and measures of social conditions in rural communities are entered as dependent variables. The independent and dependent variables are described below.

Measuring Agricultural Industrialization. Mac-Cannell uses nine variables as indicators of "agricultural industrialization" to predict social variation. These include:

• the percent of farms in the county organized as corporations farm size in acres
  
• the percent of farms in the county having more than $40,000 in sales
  
• percent of farms with full-time hired labor
  
• cost of hired labor per farm
  
• cost of contract labor per farm
  
• value of machinery per farm
  
• cost of fertilizers per farm
  
• costs of other chemicals per farm

A correlation matrix of all of these variables shows that, except for size in acres, all measures of industrialization are strongly and positively correlated, suggesting a "single, system-wide pattern of industrial agriculture." (p.37) Correlations between the nine variables and their rates of change between 1970 and 1980 are generally significant and negative, suggesting that the least industrialized of the counties underwent the most rapid change in percentage terms. Mac-Cannell suggests this trend would lead to a more uniform industrial agricultural system throughout the area.

Measuring Social Conditions in Small Communities. About 1,000 communities, averaging 5,000 population were identified within the 98-county study area. Indicators of community social conditions, the dependent variables, included the following:

• population (percent urban, rural farm and rural non-farm)
  
• median family income
  
• unemployment (percent unemployed)
  
• poverty level (percent of families and individuals living below the federal poverty standard)
  
• local government taxes
  
• per capita expenditures by local government
  
• retail sales per 1,000 persons
  
• percent employed in services
  
• percent employed in manufacture

Relationship between variables. The relationship between agricultural industrialization and community conditions was measured in several ways. Regression models were used to test: 1) the static relationship between agricultural structure in 1980 and social conditions in 1980; and 2) the dynamic relationship between agricultural structure in 1970 and the rate of change in social conditions from 1970 to 1980. All nine independent agricultural variables were used originally, but only those measures which significantly and independently predicted variation in the social variables were retained in the final analysis. The following discussion highlights a few of the relationships Mac-Cannell found.

Income. Farm size was the strongest negative predictor of median family income in the static relationship. The number of hired workers (150 + days) also significantly (and positively) predicted median family income. The only 1970 agricultural variable that predicted the rate of change in family income from 1970 to 1980 was percent farms in the county with annual sales greater than $40,000 (negative relationship).

Unemployment. Unemployment is negatively associated with both farm size and mechanization. It is positively associated with both number of hired workers (150 + days) and percent of farms w/$l00K in annual sales. MacCannell concludes from this that large, industrialized farms are major employers of unskilled workers earning low wages. He suggests that job training programs are not the answer to employment problems in these areas, but wages, working conditions and the job mix must instead be targeted for policy changes.

Poverty. In the regression analysis, 10 percent of the variation in families living below the federal poverty standard and 14 percent of the variation in individuals living below the standard is predicted by this regression model. Average farm size is the strongest, positive predictor for both variables. The pattern is even stronger when looking at the change in poverty from 1970 to 1980. "Twenty-four percent of the variation in change in individual poverty between 1970 and 1980 is accounted for by the operation of a single industrialization variable: average farm size in 1970." (p.61) The percentage of families below poverty in 1980 is also negatively predicted by the number of hired workers (150+ days). Individuals living below the poverty level in 1980 is also negatively predicted by the percent of farms corporately owned in 1980.

Farm Size Debate

The debate about the effect of farm size or scale on rural communities was first revealed in a controversial study by Walter Goldschmidt in 1944. Goldschmidt found a number of negative social effects associated with large- scale agriculture in California's Central Valley. It has been difficult for other researchers to validate Goldschmidt's findings until recently. The reason for this is that these earlier studies assumed a linear relationship between farm size and community conditions. MacCannell's study casts the farm size debate into a new light by suggesting that this relationship is curvilinear--like an inverted "J" curve--under certain assumptions about farm management and structure. He hypothesizes that community conditions remain unchanged or improve as farm size increases up to about 300 acres under conventional family management, but then decrease precipitously with further increases in size under an "industrial management" model. Goldschmidt tapped into the downward part of this curve in his study because he included one of the first communities in the Sunbelt (Arvin) in which agriculture was practiced using an "industrial management model." MacCannell suggests that his results agree with Goldschmidt's, that as the degree of agricultural industrialization increases, social conditions in the rural communities become worse.

Conclusions and Recommendations

One of the primary conclusions from this study is that the farm size debate must be reframed. "The distinction that we should be making for policy purposes is not between moderate and large-scale farming but between super- sized industrial farms...and the small- to large-sized farms...The strongest policy recommendation to be derived from these findings is the need for a policy focus on 'farms' which are much larger than those currently labeled as large farms--the super-farms." MacCannell makes the following recommendations based on this research and much of the other work of the Macrosocial Accounting Project.

• The federal government should define and move to separate super-scale farming operations from other farms which operate as small businesses
  
• The social responsibility of super-scale farms should be spelled out
  
• Special taxes on super-scale farms based on volume or profits should be considered in order to finance community development work to repair social damage
  
• The Reclamation Law should be strictly enforced
  
• Current laws against hiring illegal aliens should be enforced
  
• Minimum-wage laws for all workers should be enforced
  
• The land grant university system should be responsible for adapting and delivering all new technologies to small to moderate-scale farms
  
• Commodity programs should continue to be sales volume-sensitive up to the threshold of super-scale farms and then all support should come off.

REFERENCES

Fujimoto, Isao. 1977. The Communities of the San Joaquin Valley: The relation between scale of farming, water use, and the quality of life. Testimony before the Federal Task Force on Westlands, Sacramento, CA, August 4,1977. Reported in Community Services Task Force.

MacCannell, Dean. 1986. The Effect of Agricultural Scale on Community. Proceedings of Sustainability of California Agriculture: A Symposium.

MacCannell, Dean and Ed Dolber-Smith. 1986. Report on the Structure of Agriculture and Impact of New Technologies on Rural communities In Arizona, California, Florida, and Texas. U.S. Government Printing Office.

MacCannell, Dean and Jerry White. 1981. Agricultural land Ownership and Community Structure in California's Central Valley. Unpublished manuscript, California Policy Seminar.

MacCannell, D. and Jerry White. 1984. The social costs of large scale agriculture and the prospects for land reform in California. In Geisler, Charles C. and F. Popper (eds.) land Reform American Style. Rowan and Allanheld, Totowa, New Jersey. pp. 35-55

Macrosocial Accounting Project. 1987. Final Report on the Structure of Agriculture and Social Conditions in Rural Communities, San Joaquin Valley Drainage Study Area. Submitted to the San Joaquin Valley Drainage Program, Sacramento, CA.

For more information write to: Macrosocial Accounting Project, Dept. of Applied Behavioral Sciences, Univ. of California, Davis, CA 95616.

(GWF.O05) Contributed by Gail Feenstra